A decision methodology for large scale systems: the case of the air quality planning

The paper describes a multi-objective optimization approach to define efficientair quality control policies at regional scale. Air Quality and Cost of policy implementationare the two ingredients considered in the methodology. Both seasonal (for ozone) andyearly (for PM10) models are applied to simulate the Air Quality Index, based on ArtificialNeural Networks. Results are shown at first in terms of Air Quality models performances,and then in terms of optimal policies suggestions, over a Northern Italy domain.